Ditto: Efficient Serverless Analytics with Elastic Parallelism

Abstract

Serverless computing provides <i>fine-grained</i> resource elasticity for data analytics---a job can flexibly scale its resources for <i>each</i> stage, instead of sticking to a fixed pool of resources <i>throughout</i> its lifetime. Due to different data dependencies and different shuffling overheads caused by intra- and inter-server communication, the best degree of parallelism (DoP) for each stage varies based on runtime conditions. We present Ditto, a job scheduler for serverless analytics that leverages fine-grained resource elasticity to optimize for job completion time (JCT) and cost. The key idea of Ditto is to use a new scheduling granularity---<i>stage group</i>---to decouple parallelism configuration from function placement. Ditto bundles stages into stage groups based on their data dependencies and IO characteristics. It exploits the parallelized time characteristics of the stages to determine the parallelism configuration, and prioritizes the placement of stage groups with large shuffling traffic, so that the stages in these groups can leverage zero-copy intra-server communication for efficient shuffling. We build a system prototype of Ditto and evaluate it with a variety of benchmarking workloads. Experimental results show that Ditto outperforms existing solutions by up to 2.5&times; on JCT and up to 1.8&times; on cost.